Evaluating the impact of drought and water restrictions on agricultural production in irrigated areas through crop water productivity functions and a remote sensing-based evapotranspiration model

Agricultural Water Management 309 (2025) 109319 Available online 8 February 2025 0378-3774/© 2025 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/bync/4.0/). Evaluating the impact of drought and water restrictions on agricultural production in irrigated areas through crop water productivity functions and a remote sensing-based evapotranspiration model Joaquim Bellvert * , Magí Pamies-Sans, Jaume Casadesús, Joan Girona Efficient Use of Water in Agriculture Program, Institute of AgriFood, Research and Technology (IRTA), Parc Agrobiotech, Fruitcentre, Lleida 25003, Spain ARTICLE INFO Handling Editor Rodney Thompson Keywords: Yield Remote sensing Evapotranspiration Simulations Irrigation district Water shortage ABSTRACT In Mediterranean regions, climate change is leading to reduced precipitation, along with more frequent and severe droughts, and prolonged periods of water scarcity. In this context, as reservoir levels drop dramatically, some irrigated agricultural areas are compelled to impose water restrictions on farmers to enhance efficiency and protect crops. This study aims to evaluate the impact of varying levels of water restrictions on crop productivity across different crops, taking into account water allocation rights and the irrigation management practices of each irrigation district. Since crop yield is closely linked to the water used (crop actual evapotranspiration, ETcact), this study proposes a novel approach based on using crop water productivity functions and a remote sensing-based surface energy balance model to spatially estimate ETcact. The research was conducted across fourteen irrigation districts in Lleida and Girona, Spain, simulating six scenarios with different levels of precipitation and water rights reductions. The findings showed that reduced water availability significantly negatively affected both simulated evapotranspiration and crop yields across all districts, with variations between districts and crops. On average, yield reductions reached up to 18 % in Lleida and 16 % in Girona under the least restrictive scenarios, while more severe restrictions caused decreases of 48 % and 28 %, respectively. This approach offers valuable insights for water management agencies regarding the effects of water restrictions on crop yield losses, enabling them to make more informed decisions. Incorporating this methodology into emergency drought management plans is essential for fostering resilience in a changing climate. 1. Introduction Climate change in the Mediterranean region is projected to lead to more frequent droughts, reduced precipitation, and increased evaporation (Rocha et al., 2020). In fact, the first comprehensive scientific report on climate and environmental change in the Mediterranean concluded that this region is warming 20 % faster than the global average, with temperatures expected to rise by 2.2ºC by 2040 (MedECC, 2020). One of the most urgent challenges of the 21st century, driven by climate change, is the management and distribution of dwindling water resources. As water scarcity intensifies mostly in the Mediterranean countries, the impacts of climate change will further reduce availability for irrigation, energy production, and domestic and industrial purposes (IEMed, 2017). The primary effects of water shortage are often seen in agricultural crop production, which is expected to decline by 12–20 % by the end of the century if no adaptation measures are implemented (Lobell and Gourdji, 2012). While significant crop yield losses are more likely to occur in rainfed agriculture than in irrigated systems (Ishaque et al., 2023), the latter is, globally, at least twice as productive per unit of land (WorldBank, 2023). As a result, irrigated agriculture is anticipated to play a crucial role in ensuring global food security (Angl` es, 2011). In recent years, many regions have faced serious water security issues and shortages due to declining reservoir levels, leading to restrictions on irrigation water supplies and, consequently, significant reductions in crop production. This is particularly concerning in Catalonia, located in northeastern Spain, where low rainfall in recent years has caused reservoir water levels to drop to 42–54 % of capacity in 2022 and 21–43 % in 2023 (ACA, 2023). In response to severe drought, Catalonia activated an emergency drought management law, which outlined the rules for operating the water system and the measures that must be applied to manage the public hydraulic domain (ACA, 2020). * Corresponding author. E-mail address: [email protected] (J. Bellvert). Contents lists available at ScienceDirect Agricultural Water Management journal homepage: www.elsevier.com/locate/agwat https://doi.org/10.1016/j.agwat.2025.109319 Received 10 September 2024; Received in revised form 21 December 2024; Accepted 15 January 2025 Agricultural Water Management 309 (2025) 109319 2 This plan includes potential restrictions on water rights allocations for irrigation districts, ranging from a 25 % reduction, without distinction between crops, to allowing irrigation only for the survival of fruit trees. Water rights allocations vary widely between irrigation districts; for instance, the Algerri-Balaguer district in Lleida has an allocation of around 6000 m³ /ha, while the Garrigues Sud (GS) district only has 1300 m³ /ha. Land use, crop types, and irrigation management are also influenced to some extent by the water rights allocation of each irrigation district (Bellvert et al., 2024). When an irrigation district imposes water restrictions, the impact on crop yield will vary based on factors such as the initial allocation of water rights and its distribution throughout the season, land use, soil types, irrigation systems, and each crop’s sensitivity to water stress. In such scenarios, governments typically activate economic aid plans to compensate farmers for drought-related yield losses. For instance, in 2023, Spain received 81 million euros from the European Commission’s agricultural reserve as direct aid to support farmers facing challenges due to the lack of rainfall (La Moncloa, 2023). However, one of the major issues is the lack of consistent data on actual yield losses due to drought and different levels of water restrictions. There is currently no standardized methodology to evaluate, quantify, and forecast the impact of various water scarcity scenarios on crop yields. Having this information in advance would enable watershed policymakers and irrigation district managers to plan irrigation campaigns more effectively and develop contingency plans for different levels of water shortage. For example, they could determine how to allocate water among irrigation districts and crops based on their water requirements, sensitivity to water stress, surface area, and economic value. Yield is closely tied to the amount of water used, or evapotranspired, by plants (Doorenbos and Kassam, 1979; Steduto et al., 2012). In the late 1970s, the FAO addressed this relationship by proposing a simple equation that links relative yield reduction to the corresponding relative reduction in evapotranspiration. Therefore, any restriction on irrigation will impact crop evapotranspiration (ET) and, consequently, yield. To assess the effect of ET reductions on yield, crop water productivity functions have been established for various crops, indicating that yield responses to water deficits vary between crops (Doorenbos and Kassam, 1979; Girona et al., 2010; Greaves and Wang, 2017; L´ opez-L´ opez et al., 2018; Letseku and Grov´ e, 2022). For example, maize exhibits a linear relationship between water usage and yield (Payero et al., 2008; Letseku and Grov´ e, 2022). Conversely, Naor and Girona (2012) found that crop evapotranspiration in apple trees can be reduced by about 15–20% of its potential water needs without negatively affecting yield. Yield forecasting plays a crucial role in effective risk management for various stakeholders, including farmers, insurers, reinsurers, and governments (Li et al., 2021). Over the years, numerous approaches have been employed to forecast yield, each offering different levels of granularity, accuracy, and timing. Among these, crop growth models have been widely used to simulate crop yield (Stockle and Nelson, 1994; Keating et al., 2003; Jones et al., 2003; Hunt et al., 2006; Steduto et al., 2009). However, in some regions, the performance of these models at regional scale is limited by the lack and uncertainty of available input data, such as detailed soil or irrigation type maps (Bouman, 1995; Lüke and Hack, 2017). Recent advancements in remote sensing, especially at high spatial and temporal resolutions, have enabled the development of innovative methods to estimate vegetation biophysical variables and crop evapotranspiration. These advancements help reduce some of the uncertainties associated with yield estimations. Some methods estimate yield using empirical regressions with spectral vegetation indices alone (Teal et al., 2006; Huang et al., 2014; Li et al., 2022), or in combination with other data, such as meteorological information (Dabrowska-Zielinska et al., 2002; Kern et al., 2018). Others predict yield by assimilating remote sensing observations into crop or agrohydrological models (Prasad et al., 2006; Ines et al., 2013; Vazifedoust et al., 2009; Li et al., 2014). However, few studies have focused on using remote sensing for evapotranspiration (ET) retrieval to estimate yield at regional level. Among these, several studies have employed an evaporative stress index (ESI) to assess water stress and predict yield, often achieving higher correlations than with vegetation indices alone (Anderson et al., 2016; Yang et al., 2018, 2021). Over the past two decades, remote sensing surface energy balance (SEB) models have made significant strides, offering diagnostic assessments of ET at very high spatio-temporal resolutions. Numerous SEB schemes have been developed, varying in complexity (Norman et al., 1995; Anderson et al., 1997, 2004; Bastiaanssen et al., 1998; Allen et al., 2007; Boulet et al., 2015). Among these, the two-source energy balance (TSEB) model developed by Norman et al. (1995) is one of the most widely used and robust. This has the advantadge that partitions surface fluxes between soil and canopy components. Recently, the TSEB modeling scheme has been used with Copernicus-based inputs and data mining approaches to derive energy fluxes at a 20-meter resolution (Guzinski et al., 2021; Jofre-ˇ Cekalovi´ c et al., 2022). Previous studies have demonstrated the effectiveness of this approach in assessing differences in crop water requirements across several irrigation districts in the Ebro basin (Bellvert et al., 2024). Despite recent advances in yield forecasting, most studies have focused on annual crops. The complexity of predicting yield in woody crops has resulted in a lack of research in this area. Similarly, there is lack of district-level crop yield simulations under varying levels of irrigation water restrictions due to water shortages. To address this gap, this manuscript proposes a novel approach using the remote sensing TSEB ET-based scheme to simulate the impact that diferent water restrictions levels, applied in irrigation districts with varying water allocations, may have on the production of various crops. The study was conducted in two distinct irrigated areas in Catalonia (Lleida and Girona), each with different agrometeorological conditions, crop type distribution, and irrigation districts with varying water right allocations. 2. Materials and methods 2.1. Study sites The study area encompasses the regions of Lleida and Girona in Catalonia, Spain, situated in the northeastern part of the Iberian Peninsula (Fig. 1). These two regions are the most significant for irrigated agriculture in Catalonia and were the hardest hit by water restrictions during the 2023–2024 drought. The Lleida area is part of the Ebro basin, while the selected area in Girona lies within the Low Ter river basin. The irrigation districts (IDs) in each basin are detailed in Table 1. In the Ebro basin of Lleida, eight IDs were analyzed, covering a total irrigated area of 150,281 has. For the GS district, 50 % of its area was included in the analysis. In the Low Ter basin of Girona, five IDs were examined, with a combined irrigated area of 13,633 has. Water right allocations, irrigation systems and level of modernization varied across the IDs. In Lleida, the least modernized district is Canals d’Urgell, with nearly 65,000 has under surface irrigation. Nevertheless, the most modernized is Segarra-Garrigues, with an irrigated area in 2023 of around 25000 ha. Both regions experience a typical Mediterranean climate, characterized by mild, wet winters and hot, dry summers (IPCC, 2022). However, notable differences exist in rainfall between the two areas, which means that Lleida has a semi-arid climate, while Girona is humid. Over the five years analyzed, Lleida recorded an average annual precipitation of 384 mm and a reference evapotranspiration (ETo) of 1174 mm, while Girona registered 576 mm of precipitation and 1044 mm of ETo. Precipitation records were obtained from the Official Meteorological Service of Catalonia (SMC) (www.meteo.cat). The SMC developed a product called the Integrated Hydrometeorological Tool (EHIMI), which generates precipitation maps at a 1 km resolution. These maps are created using advanced radar image processing techniques and algorithms that improve the accuracy of meteorological radar estimates. On the other hand, ETo was calculated using the FAO-56 J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 3 Penman-Monteith method (Allen et al., 1998) and meteorological data from the ERA-5 Land reanalysis dataset of the Copernicus Climate Change Service (Mu˜ noz Sabater, 2019).The crop distribution is similar in both regions, with the predominant crops, in descending order of irrigated area, being wheat/barley, maize, olive, alfalfa, peaches, almond, pear, apple, grapevine, festuca, lolium, apricot, and pistachio (Bellvert et al., 2024). Collectively, these crops account for more than 96 % of the irrigated area. 2.2. Remote sensing for evapotranspiration Crop actual evapotranspiration (ETcact) of both areas of interest was computed during years 2018–2022. The Two-source energy balance (TSEB) modeling scheme using Copernicus-based input was used in this study due to its high accuracy in estimating ETcact (Guzinski et al., 2021; Jofre-ˇ Cekalovi´ c et al., 2022; Bellvert et al., 2024). These studies reported a RMSE of instantaneous latent heat flux of around 30 % in agricultural areas. The TSEB was first described by Norman et al. (1995) with important adjustments described in Kustas and Norman (1999). Fig. 1. Study areas in Lleida (red box) and Girona (green box), displaying the mean cumulative actual crop evapotranspiration (ETcact) for the years 2018–2022, estimated through remote sensing for both regions and for all irrigated plots simulated within each irrigation district. Table 1 Summary of the total irrigated area, annual water rights allocations, and the percentage of each irrigation system within each irrigation district (ID) of the Lleida Ebro and Girona Low Ter basins. ID Name Irrigated area (ha) Water rights allocation (m 3 /ha) Irrigation system (%) Drip Sprinkler Surface Lleida - Ebro basin      AB Algerri-Balaguer 6548 ~6000 22.3 76.1 1.6 C Carrassumada 1346 ~7000 84.1 14.1 1.8 CAYC Canal Arag´ on y Catalu˜ na 29635 ~8000 33.0 65.1 1.9 CP Canal de Pinyana 9223 ~10000 35.1 40.2 24.7 CU Canals d’Urgell 64333 ~9000 9.9 5.4 84.7 GS Garrigues Sud 7365 ~1300 100 - - SG Canal Segarra-Garrigues 25383 1500–6500 a 82.5 15.4 2.1 SS Segri` a Sud 6448 ~2000 94.4 0.1 5.5 Girona - Low Ter basin      BF Baix Fluvi` a 4531 ~8500 28.1 3 68.9 CT Cervi` a de Ter 522 ~8667 0.3 0.2 99.5 MME Muga marge esquerra 2631 ~5500 5.1 1.2 93.7 MP Molí de Pals 2823 ~9970 13.7 2.4 83.9 PC Presa de Colomers 2849 ~8000 18.8 0.8 80.9 SV S` equia de Vinyals 277 ~8000 0.4 - 99.6 a Simulations were conducted considering a water right allocation of 4500 m 3 /ha J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 4 The Priestley-Taylor (Priestley and Taylor, 1972) iterative retrieval was used in order to partition directional radiometric temperature, T rad (ɵ), into soil and canopy temperature (T s and T c , respectively), based on the vegetation cover fraction at the thermal sensor view angle, f c (ɵ). Bellvert et al. (2024) describes in detail the methodology used. The Copernicus-based inputs used were the following: Sentinel-2, and Sentinel-3 imagery, meteorological data from the European Center of Medium Weather Forecast (ECMWF), and Land Cover map from the Copernicus Climate Change Service (C3S). In this study, four tiles from Sentinel-2 were downloaded in Lleida (T31[TBF, TCG, TBG, TCF]) and two in Girona (T31[TDG, TEG]). In Lleida, a total of 35, 46, 45, 40, 35 and 44 Sentinel-2 cloud-free images were downloaded for the period 15th March to 30th October respectively for 2017–2022. On the other hand, 104, 101, 227, 228, 229 and 230 Sentinel-3 cloud-free scenes were respectively fetched for the same years. On the other hand, in Girona, a total of 39, 40, 46, 38 and 44 Sentinel-2 cloud-free images were downloaded for the period 15th March to 30th October respectively for 2018–2022. On the other hand, 172, 227, 230, 196 and 230 Sentinel-3 cloud-free scenes were respectively fetched for the same years. Biophysical variables of the vegetation were computed at 20 m spatial resolution from Sentinel-2 using the Biophysical Processor available in the SNAP software v9.0 (https://step.esa.int/main/download/snap-download/—last accessed 20.07.2024). The obtained biophysical parameters were Lea Area Index (LAI), fraction of vegetation cover (FVC), Fraction of Absorbed Photosynthetically Active radiation (FAPAR), Canopy Chlorophyll Content (CCC) and Canopy Water Content (CWC). Python scripts were then used to estimate the fraction of vegetation that is green (fg – green LAI over total LAI), vegetation gap fraction observed at the sensor viewing angle (fc (θ)), and leaf bihemispherical reflectance and transmittance, together with constant values for soil reflectance in the VIS-NIR to quantify the shortwave net radiation of the soil and canopy (F´ eret et al., 2017). On the other hand, the thermal data needed to drive the ETa was obtained from the SLSTR sensor on board of the Sentinel-3 satellite. Thermal imagery is at 1 km spatial resolution and less than two days’ temporal resolution. A thermal Data Mining Sharpening (DMS) approach (Gao et al., 2012) was used to sharpen the 1 Km Sentinel-3 land surface temperature at 20 m resolution. Each Sentinel-3 scene was matched with a Sentinel-2 scene acquired at most ten days before or after the Sentinel-3 acquisition. A regression model based on bagging ensemble of decision trees was used to relate LST to a suite of shortwave spectral reflectance from Sentinel-2 and ancillary data, at coarse spatial resolution, and then to apply it to fine pixel resolution (Guzinski et al., 2020). The output of the sharpening was a 20 m representation of the LST. More information about the approach and code used is available online (https://github.com/radosuav/pyDMS, last accessed: 10 February 2023). Meteorological data used in this study was obtained from the ERA-5 Land reanalysis dataset of the Copernicus Climate Change Service (Hersbach et al., 2020) and from the Copernicus Atmosphere Monitoring Service (CAMS). The data used from ERA5 consists of air temperature and dew point temperature at 2 m, wind speed at 100 m, surface pressure, total column water vapour (TCWV) and surface geopotential. Aerosol optical thickness (AOT) at 550 nm was obtained from CAMS since it is not included in ERA5. A digital elevation model (DMS) elevation model was used to enhance the spatial resolution of air temperature, vapor pressure, and surface pressure and to correct its variations due to changes in elevation at blending height. The final instantaneous parameter was surface solar irradiance. Clear sky conditions were assumed since this parameter is only used at a time and place where thermal observations of the surface by the Sentinel-3 satellite were possible. In addition, solar irradiance was firstly estimated using AOT, TCWV and surface pressure and subsequently corrected by elevation, incidence angle, and terrain shading (Guzinski et al., 2021). Ancillary data related to structural vegetation variables and leaf inclination angle were determined from a look-up table associated to a crop class according to Guzinski et al. (2020). The crop class maps used in this case was the one from CS3 landcover map, produced at 300 m resolution (https://cds.climate.copernicus.eu/cdsapp#!/dataset/satellite-land-cover?tab=overview, last accessed: 10.10.2022). Once ETcact was obtained for all cloud-free dates, a gap filling approach was applied based on using a crop stress coefficient (K cs ) on the assumption that the ratio of reference (FAO-56) to actual ET remains steady over short periods. The detailed methodology is described by Bellvert et al. (2024), which used the same time-series of ETa for the study area of Lleida during the period 2017–2022. 2.3. Crop type classification An administrative database, SIGPAC-DUN, was used in order to know the crop type of each plot during the growing seasons 2018–2022. SIGPAC-DUN contains information about crop type of the plots in Spain, reported by farmers (as mandated by the European crop subsidy program) following the single agrarian declaration (Declaraci´ o Única Agr` aria, DUN) and provided annually by the Spanish Agricultural Land Geographic Information System (SIGPAC) public administrative database. Annual SIGPAC-DUN products are obtained from: https:// agricultura.gencat.cat/ca/ambits/desenvolupament-rural/sigpac/ index.html. Since 2021, SIGPAC-DUN also contains information about the irrigation system and double cropping. However, Paolini et al. (2022), found some discrepancies between the SIGPAC-DUN map of Lleida and the one they obtained using artificial intelligent algorithms and remote sensing data. Thus, in the Lleida region, we used the irrigation system maps obtained by Paolini et al. (2022), while for Girona, the SIGPAC-DUN irrigation system maps were used. 2.4. Crop water productivity functions Since crop yield is directly linked to the amount of water used (evapotranspiration) (Doorenbos and Kassam, 1979; Hanks, 1974), a comprehensive review of crop water productivity functions was carried out (Steduto et al., 2012). The main crops from both study areas, accounting for 97 % of the cultivated land, were selected for analysis. Table 2 presents the water productivity functions used in this study, along with the corresponding potential yield values for each crop. These potential yield values, representing the maximum attainable yield under fully irrigated conditions, were obtained through surveys conducted with leading cooperatives, technicians, and farmers in each region. 2.5. Simulations of yield in the different scenarios Fig. 2 provides an overview of the methodology used in this study. Simulations were conducted separately for the Lleida and Girona regions. The cumulative actual crop evapotranspiration (ETcact) for each year, estimated using remote sensing, was used as the reference crop evapotranspiration for a potential year, denoted as ETcref act. The total water available for the crop (CWA) was then estimated for each plot as follows: CWA =P x Peff + (WA −Ex)x Ieff (1) where, CWA corresponds to crop water availability (mm), P is the annual precipitation; Peff is effective precipitation, which was calculated as half of the precipitation for single events with more than 10 mm, and otherwise was considered to be zero (Olivo et al., 2009); WA corresponds to water allocation of the irrigation district (m 3 /ha) (Table 1); Ex is the excess of water, which it is understood as the amount of water that a grower has available to irrigate, but not used. It depends, in part, on the efficiency of the irrigation system and distribution, but also on the technology level of the grower to irrigate based on the crop water requirements. The default minim values of Ex considered for each irrigation system were 100, 50 and 0 mm for drip, sprinkler and flood irrigation, respectively. Finally, Ieff values which correspond to the J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 5 irrigation system efficiency were 0.9, 0.75 and 0.60 for drip, sprinkler and flood irrigation, respectively. For a reference year, CWA and its components were adjusted to each district and cropping scenario, as follows. Since CWA of each plot in potential conditions should be similar to ETcref act, the following iterative methodology was used to adjust it: If ratio =CWA ETcref act <1 then, CWA =ETcref act If ratio =CWA ETcref act >1 then, the following methodology was carried out: firstly, Ex of drip and sprinkler plots was modified adding intervals of 50 mm until the ratio was ≤1. Ex of flood irrigated plots was not considered. The iterative process was conducted until a maximum of eight times. After this first step, for those plots with still a ratio >1, the parameter Ieff was modified. The default values of Ieff were decreasing at intervals of 0.02 until that the ratio was ≤1. This iterative process was conducted a maximum of twenty times. Finally, the CWA of the remaining plots with still a ratio >1, but closer to one, were considered equal to ETcref act. On the other hand, simulations of actual crop evapotranspiration (ETcsim act ) were also obtained at plot level in different scenarios of water restrictions. To achieve it, ETcsim act was considered to be the same as the simulated crop water availability (CWAsim).The latter was simulated into five contrasting scenarios, which consisted on reducing different percentages of precipitation and irrigation district water rights allocation (Table 3). Therefore, in each of these scenarios, the parameters water rights allocation (WA) and annual precipitation (P) of Eq. 1 were modified by reducing them at different percentages. The total amounts of P and WA simulated for each scenario and irrigation district are shown in Fig. 3. To deal with scenarios of restrictions in water availability, the relative yearly crop water availability (CWArel)of each plot was defined as: CWArel =CWAsim CWA (2) In order to estimate the relative yield (Yieldrel)for each plot under each simulated scenario, the CWArel value was used as the dependent variable (x) in the water productivity functions defined in Table 2. Then, actual yield for each simulated scenario (Yieldsim)was obtained as: Yieldsim =Yieldrel ×Yieldpot field (3) where Yieldpot field is defined as: Yieldpot field =CWAmean ×Yieldpot CWAmean%maxALL (4) This correction of potential yield (Yieldpot field)for each plot is necessary due to differences in water allocations, not all the irrigation Table 2 List of water productivity functions used for each crop. Crop Functions Threshold a b c Reference Potential yield (tonnes/ha) Apple y =ax+b 0.78 1.47 −0.14 Steduto et al. (2012) 60 Almond y =ax+b 0.87 1.17 −0.02 2.5 Pear y =ax+b 1 0.94 0.06 35 Peach y =ax+b 1 1.13 −0.13 35 Olive y =ax 2 +bx+c 1 −2.55 5.19 −1.64 10 Grapevine y =ax 2 +bx+c 1 −3.15 6.06 −1.91 10 Pistachio y =ax+b 0.71 1.89 −0.55 6 Apricot y =ax+b    15 Annual crops y=ax+b - 1 0   Maize 14 Wheat 8 Alfalfa 30 Festuca 9 Barley 11 Lolium 7 Fig. 2. Flowchart of the methodology used in the study. Table 3 Simulated scenarios with varying percentages of reductions in water rights allocations and precipitation. Name Description Water rights allocations (%) Precipitation (P) Control 100 % irrigation water right allocation 100 100 Pr75 25 % reduction of precipitation 100 75 Pr50 50 % reduction of precipitation 100 50 Irri75 25 % reduction of irrigation water right allocation 75 100 Irri50 50 % reduction of irrigation water right allocation 50 100 Irri25 75 % reduction of irrigation water right allocation 25 100 J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 6 districts would have the same potential yield. Yieldpot corresponds to potential yield values of Table 2. CWAmean corresponds to the mean CWA value of each plot. CWAmean%maxALL corresponds to the highest CWA mean value for each irrigation district and crop type. Finally, in order to calculate the total yield for each crop and irrigation district, Yieldsim was multiplied by the total number of has: Yieldtotal =Yieldsim ×Area of ID (5) The regression between simulations of actual crop evapotranspiration (ETcsim act ) in each contrasting scenario and yield (Yieldsim)were compared among irrigation districts in relative values. To achieve it, both parameters were normalized for each crop type as: ETcsimREL act =CWAsim −CWAsim_MIN CWAsim_MAX −CWAsim_MIN (6) YieldsimREL =Yieldsim −Yieldsim_MIN Yieldsim_MAX −Yieldsim_MIN (7) where the subindices sim_MAX and sim_MIN respectively corresponds to maximum and minimum values of each parameter within each crop type. 3. Results 3.1. Statistical analysis across irrigation districts and areas of study In the current scenario of irrigation districts (IDs) without water restrictions, the statistical analysis revealed significant differences in both ETcref act and Yieldsim across crop types, irrigation districts, and study areas (p <.0001). Additionally, the interactions between these factors were also significant in both regions. Notable differences were also observed in precipitation (P) and modeled irrigation efficiency (Ieff ) between the irrigation districts and study areas (Fig. 4). Overall, the irrigation efficiency (Ieff ) in Girona was significantly lower than in Lleida, with CU and SV exhibiting the lowest values in Lleida and Girona, respectively. This could be explained by a higher percentatge of land irrigated by flooding. Fig. 4 shows that ETcref actand CWA have similar values, indicating the strong performance of the iterative methodology applied. When comparing irrigation districts with full water allocation rights (Control), both parameters were significantly higher in Lleida than in Girona (Figs. 4 and 5). This is related to the fact that Lleida has a higher ETo than Girona. In Lleida, the irrigation districts AB, CAYC, CU, C, and CP exhibited the highest average cumulative CWA or ETcref act, with no significant differences among them, followed by SG, GS, and SS, which had significantly lower values. A similar pattern was observed for average Yieldsim, with GS and SS showing the lowest yields. Overall, the average yield for GS and SS under the Control scenario was 65 % lower compared to districts with full water rights allocation. In Girona, MME was the only district with significantly lower CWA rates, and in terms of Yieldsim, MME, CT, and SV had the lowest values. 3.2. Impact of simulations on ETsim a and Yieldsim across regions and irrigation districts This analysis compared each scenario with the potential ETcsim act and Yieldsim rates within each irrigation district. Overall, a reduction in water contributions, whether due to decreased precipitation or irrigation water rights allocation, had a significant negative impact on both crop Fig. 3. Amounts of precipitation and water rights allocations simulated for various scenarios in each irrigation district of Lleida (left) and Girona (right). J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 7 evapotranspiration and yield (Figs. 5 and 6). The decline in ETcsim act under the Pr75, Pr50, and Irri75 scenarios was similar across all irrigation districts in Lleida and Girona with water rights exceeding 5000 m³ /ha (Table 1). On average, ETcsim act decreased by 19 % compared to the Control scenario (Fig. 6). However, the reduction in ETcsim act for the more restrictive scenarios (Irri50 and Irri25) was significantly more pronounced in the irrigation districts of Lleida compared to those of Girona. Specifically, the reduction was 28 % and 46 % in Lleida, compared to 18 % and 29 % in Girona, respectively, for the Irri50 and Irri25 scenarios. In terms of Yieldsim, the average reduction in Lleida for the Pr75, Pr50, Irri75, and Irri50 scenarios was 18 %, while in Girona, it was 16 %. Additionally, the Irri25 scenario resulted in yield reductions of 48 % in Lleida and 28 % in Girona. The two irrigation districts with water allocations below 5000 m³ / ha (GS and SS) exhibited significantly lower ETcsim act and Yieldsim values under the Control scenario, and consequently in all other simulated scenarios as well (Fig. 5). Their response to varying levels of water restrictions and reduced precipitation was more pronounced compared to other districts. Both ETcsim act and Yieldsim declined more in scenarios with reduced precipitation than in those with restricted irrigation water. For example, the average ETcsim act and Yieldsim for GS and SS decreased by 37 % and 53 % in the Pr75 and Pr50 scenarios, respectively. In contrast, the reduction was 33 % in the Irri75 and Irri50 scenarios, and 46 % in the Irri25 scenario (Figs. 5 and 6). In contrast, the irrigation districts in Girona showed a less pronounced response to more severe water restrictions (Irri50 and Irri25) compared to those in Lleida. In these two scenarios, ETcsim act and Yieldsim decreased by 17 % and 27 %, respectively. Regarding water productivity (WP), calculated from the ratio between ETcsim act and Yieldsim, no significant differences were observed between simulated scenarios in any of the IDs in Lleida or Girona (Fig. 5). The wide variability in yield values across different crops likely obscured the statistical analysis, making it necessary to analyze the data separately for each crop. 3.3. Impact of simulations on Yieldsim across different crops Because each crop responds differently to water deficits, it was necessary to analyze the effects of various water restriction scenarios individually for each crop. First, it is important to note that there were significant differences in ETcref act between crops. Overall, alfalfa, apple, pear, peach, and maize exhibited the highest evapotranspiration rates, while barley, wheat, festuca, and pistachio showed the lowest (Figs. 7 and 8). The lowest ETcref actrates for pistachio could be explained by the young age of most plantations in the area. ETcref act rates were significantly higher in Lleida than in Girona, partly due to the higher reference evapotranspiration (ETo) in Lleida. For all crops, reductions in irrigation water allocations (Irri75, Irri50, and Irri25) had a greater impact on Yieldsim in Lleida’s irrigation districts compared to those in Girona (Figs. 7 and 8, Tables 4 and 5). Specific details about Yieldsim reductions in Lleida are presented in Fig. 7 and Table 4, where a significant interaction between crops and irrigation Fig. 4. Descriptive statistical analysis of differences in precipitation (P), irrigation efficiency (Ieff ),), reference actual evapotranspiration (ETcref act), modeled crop water availability (CWA), and yield (Yieldsim) between irrigation districts in Lleida and Girona under the current full water allocation scenario (Control). Different letters mean significant differences at p-value. J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 8 Fig. 5. Mean values of cumulative simulated average actual crop evapotranspiration (ETcsim act ), average yield (Yieldsim), and water productivity (WP) for each irrigation district in Lleida (left) and Girona (right) across each simulated scenario (Control, Pr75, Pr50, Irri75, Irri50, Irri25). Different letters mean significant differences between irrigation districts at p ≤0.05 using Tukey’s honest significant difference test. Fig. 6. Differences in the reduction of simulated actual crop evapotranspiration (ETcsim act ) and yield (Yieldsim) between irrigation districts in Lleida (a, b) and Girona (c, d) for each simulated scenario (Pr75, Pr50, Irri75, Irri50, and Irri25) compared to the Control. Scenarios marked with an asterisk (*) indicate significant differences at p <0.05. J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 9 districts is evident. The most substantial reductions in Yieldsim under the Pr75 and Pr50 scenarios were observed in the GS and SS irrigation districts, which had the lowest water allocations and, consequently, the lowest ETcsim act rates (Fig. 5). These two irrigation districts are thus highly dependent on green water to ensure crop survival. Crops in these two irrigation districts, including olives, grapevines, peaches, and apricots, experienced the most significant reductions in Yieldsim under the Pr75 and Pr50 scenarios. Olive trees in GS saw yield decreases of up to 76.8 %, producing only around 1.7 tonnes per ha compared to the potential yield of 10 tonnes per ha. Similarly, these crops were more sensitive to yield declines under irrigation water restriction scenarios (Irri75, Irri50, and Irri25), with yield decreasing as water restrictions became more severe. This pattern was consistent across all irrigation districts, with the most pronounced yield decreases observed again in GS and SS. However, it is important to note that the potential yield of these two IDs was already significantly lower than others, so the higher relative percentage decreases might translate to smaller absolute values compared to other IDs. Conversely and in the context of this study, apple trees demonstrated the lowest sensitivity to yield decline under different scenarios. While no significant differences in Yieldsim reductions were observed among the group of annual crops -such as maize, wheat, alfalfa, festuca, barley, and lolium-, alfalfa appeared to be slightly more sensitive than the others. The Yieldsim reductions in Girona are presented in Fig. 8 and Table 5. Overall, the simulations followed a pattern similar to that observed in Lleida. Specifically, apple and almond, followed by grapevines and olives, experienced less pronounced yield reductions under the Pr75, Pr50, Irri75, and Irri50 scenarios. However, under the scenario of severe irrigation water restrictions (Irri25), these crops exhibited a similar response to other annual crops such as maize, wheat, festuca, barley, and lolium. As in Lleida, yield reductions for alfalfa were significantly more pronounced compared to other crops. The yield response of each crop to water deficits was defined by the water productivity functions shown in Table 2. Although the water production functions of the major herbaceous crops are essentially lineal, those of some woody crops are asymptotical. Figs. 9 and 10 illustrate these relative functions for Lleida and Girona, respectively, using data obtained from simulations in each irrigation district and crop between 2018 and 2022. It is interesting to observe the varying shapes of these functions depending on the crop type, and how the relationship between ETcsimREL act and YieldsimREL declines as water restrictions increase. Fig. 7. Response of cumulative simulated actual crop evapotranspiration (ETcsim act ) and yield (Yieldsim) under different scenarios (Control, Pr75, Pr50, Irri75, Irri50, Irri25) for each crop and irrigation district in Lleida. J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 16 decreased. The variability observed between different studies and our simulations can be attributed to differences in irrigation systems and management practices, soil types, and the phenological stages at which maximum water deficit levels occur. Many researchers have noted that the yield response to water stress primarily depends on its intensity and the phenological stage during which it occurs (Li et al., 1989; Romero et al., 2004; Basile et al., 2011). It is well-known that climate change is likely to negatively impact Fig. 12. Maps showing yield (Yieldsim) (left column), cumulative actual crop evapotranspiration (ETcsim act ) (central column), and water productivity (WP) (right column) under the simulated scenarios Control, Pr75, Pr50, Irri75, Irri50, and Irri25 for the irrigated plots in Girona. J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 17 irrigated agriculture in southern Europe. However, the severity of these impacts will depend on policy choices made at both the farm and water body agency levels. In cases of severe water shortages, the information provided by this research could aid in establishing a fair distribution of water restrictions among irrigation districts (IDs) and in assessing their economic impact. Lorite et al. (2007) suggested that when water allocations fall significantly below the crop’s water requirements for optimal yield, a suitable strategy is to allocate water to crops with high water productivity (primarily fruit trees) and adjust the cropping pattern accordingly. Potential actions include varying water restrictions among IDs based on their impact on yield losses, reallocating water to areas with a higher proportion of drought-resistant crops, prioritizing water for high-value crops (although this may be contentious), and promoting regulated deficit irrigation (RDI) strategies where feasible. However, only the most technically advanced farmers and modern irrigation districts with pressurized systems will be able to implement RDI (Girona et al., 2003). Additional measures, some of which were already adopted during the 2023 drought, could involve restricting irrigation to woody crops to ensure their survival, irrigating only 50 % of the surface area, prioritizing fields with efficient irrigation systems, or imposing water restrictions proportional to the crop’s water requirements. Simulations carried out in this research were obtained using crop type maps from years 2018–2022 and adopting water restrictions equally for all crops within each irrigation district. In Catalonia (Spain), these maps are available in open-access through SIGPAC, but they tipycally become accessible only at the end of the growing season. This delay can be a limitation when the goal is to obtain simulations to make real-time decisions for the same growing season. In addition to this, it is important to note that these maps are not available for all regions. Therefore, the use of remote sensing technologies capable of identifying crop types early in the growing season can be highly advantageous for water management, particularly for optimizing water Distribution among crops and irrigation districts. Solutions based on the combined use of deep learning architectures with Sentinel-2 time series have proven effective for this purpose (Rußwurm and K¨ orner, 2018, Gen- ´ e-Mola et al., 2024). On the other hand, based on recent experiences, it is evident that during periods of water restrictions caused by drought, farmers often adapt by changing their practices; for example, they might cultivate entirely different crops or even decide not to sow anything at all. This is especially common in summer crops or double cropping. Water requirements of these crops are very high and under these situations of water shortage, farmers usually decide to replace them only for winter cereals or less water demanding crops such as sunflower. Future refinements of this approach should focus on enhancing all the previously mentioned shortcomings, while also considering the carry-over effects of drought on yield in consecutive years and conducting more simulations that account for combined reductions in both precipitation and irrigation. Finally, although this study focuses solely on quantifying yield losses under different water restriction scenarios, future research should address translating these losses into economic terms. Estimating economic losses due to drought is a contentious issue, with values varying significantly depending on the assessing entity and the methodology used. The approach developed in this study aims to contribute to the establishment of a standardized methodology for simulating crop production losses in a specific region. For example, under the Irri25 scenario, the total yield loss for maize—the predominant summer crop in both regions—would amount to 685,000 tonnes and 103,000 tonnes in Lleida and Girona, respectively (Table 6). Given that the average market price in recent years was 240 euros per tonne, the total economic loss could exceed 189 million euros for this crop alone. 5. Conclusion Drought, marked by insufficient precipitation, limits water resources, forcing irrigation communities to restrict usage, which reduces agricultural output and causes economic hardships for farmers and the wider community. In response to these challenges, this study has developed an innovative methodology based on remote sensing for crop evapotranspiration, capable to evaluate the impact of various irrigation water restrictions on production with a focus on differentiating between irrigation districts and crop types. The results demonstrated that reducing irrigation water allocations significantly negatively impacts crop evapotranspiration and, consequently, yields. Significant differences were observed across both irrigation districts and crops. Irrigation Table 6 Total yield amounts (in tonnes ×10 −3 ) for each crop under each simulated scenario (Control, Pr75, Pr50, Irri75, Irri50, and Irri25) in the irrigated areas of Lleida and Girona. J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 18 districts with low water rights allocation, such as GS and SS, were found to be highly dependent on precipitation. The most drastic reductions in yield were observed under the Pr50 scenario for grapevines, peaches, apricots, and olives. Conversely, irrigation districts with high water allocations experienced production reductions of up to 48 % in Lleida and 28 % in Girona. This approach, proven to be both practical and adaptable for global use, is expected to benefit irrigation district managers and policymakers and could be integrated into emergency drought management plans to enable more informed decision-making. However, there is still room for improvement, and future research should focus on incorporating seasonal crop sensitivity to water stress, validating yield values, and achieving a more precise distribution of crop and soil types. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper Acknowledgements This research was supported by the Department of Climate Action, Food and Rural Agenda (DACC) of the government of Catalonia, under the framework of drought during 2023. In addition, projects ET4DROUGHT (No. PID2021-127345OR-C31) and DIGISPAC (TED2021-131237B-C21) both funded by the Spanish Research Agency of the Ministry of Science and Innovation (MICINN-AEI) of Spain, also contributed with this research. We also extend our gratitude to the Servei Meteorol` ogic de Catalunya (SMC), particularly Vicent Altava, for providing us with the high-resolution precipitation maps. References ACA., 2020. ACORD GOV/1/2020, de 8 de gener, pel qual s ′ aprova el Pla especial d ′ actuaci´ o en situaci´ o d ′ alerta i eventual sequera. (gencat.cat) (last accessed August 18, 2023). ACA., 2023. Portal de la sequera (gencat.cat) Estat de les reserves d ′ aigua als embassaments. El portal de la sequera (gencat.cat) (last accessed February 18, 2024). Allakonon, M.G., Akponikp` e, P.B., 2022. Relationship of maize yield to climatic and environmental factors under deficit irrigation: a quantitative review. Int. J. Agron. 2408439. https://doi.org/10.1155/2022/2408439. Allen, R.G., Pereira, L., Raes, D., Smith M., 1998. Crop evapotranspiration: guidelines for Computing Crop Water requirements. FAO Irrigation and Drainage Paper 56. Food and Agriculture Organisation, Rome, Italy. Allen, R.G., Tasumi, M., Trezza, R., 2007. Satellite-based energy balance for mapping evapotranspiration with internalized calibration (METRIC)—Model. J. Irrig. Drain. Eng. Asce 133, 380–394. Anderson, M.C., Hain, C.R., Jurecka, F., Trnka, M., Hlavinka, P., Dulaney, W., Otkin, J. A., Johnson, D., Gao, F., 2016. Relationships between the evaporative stress index and winter wheat and spring barley yield anomalies in the Czech Republic. Clim. Res. 70 (2/3), 215–230. https://doi.org/10.3354/CR01411. Anderson, M.C., Norman, J.M., Diak, G.R., Kustas, W.P., Mecikalski, J.R., 1997. A twosource time-integrated model for estimating surface fluxes using thermal infrared remote sensing. Remote Sens. Environ. 60, 195–216. https://doi.org/10.1016/ s0034-4257(96)00215-5. Anderson, M.C., Norman, J.M., Mecikalski, J.R., Torn, R.D., Kustas, W.P., Basara, J.B., 2004. A multiscale remote sensing model for disaggregating regional fluxes to micrometeorological scales. J. Hydrometeorol. 5, 343–363. https://doi.org/ 10.1175/1525-7541(2004)005<0343:AMRSMF>2.0.CO;2. Angl` es., 2011. La demanda d’aigua per a la producci´ on d’aliments a Catalunya. D` eficit actual i perspectives de futur. Estudis d’hist` oria agraria 23, 107-115. ISSN 02104830. Basile, B., Marsal, J., Mata, M., Vallverdú, X., Bellvert, J., Girona, J., 2011. Phenological sensitivity of Cabernet Sauvignon to water stress: vine physiology and Berry composition. Am. J. Enol. Vitic. 62, 452–461. https://doi.org/10.5344/ ajev.2011.11003. Bastiaanssen, W.G.M., Menenti, M., Feddes, R.A., Holtslag, A.A.M., 1998. A remote sensing surface energy balance algorithm for land (SEBAL). 1. Formulation. J. Hydrol. 212–213, 198–212. https://doi.org/10.1016/S0022-1694(98)00253-4. Bellvert, J., Pamies-Sans, M., Quintana-Seguí, P., Casadesús, J., 2024. Analysis and forecast of crop water demand in irrigation districts across the eastern part of the Ebro river basin (Catalonia, Spain): estimation of evapotranspiration through Copernicus-based inputs. Irrig. Sci. https://doi.org/10.1007/s00271-024-00971-1. Boulet, G., Mougenot, B., Lhomme, J.-P., Fanise, P., Lili-Chabaane, Z., Olioso, A., Bahir, M., Rivalland, V., Jarlan, L., Merlin, O., et al., 2015. The SPARSE model for the prediction of water stress and evapotranspiration components from thermal infra-red data and its evaluation over irrigated and rainfed wheat. Hydrol. Earth Syst. Sci. 19, 4653–4672. https://doi.org/10.5194/hess-19-4653-2015. Bouman, B.A.M., 1995. Crop modelling and remote sensing for yield prediction. Neth. J. Agric. Sci. 43, 143–161. https://doi.org/10.18174/njas.v43i2.573. Cavero, J., Faci, J.M., Medina, E.T., Martínez-Cob, A., 2017. Alfalfa forage production under solid-set sprinkler irrigation in a semiarid climate. Agric. Water Manag. 191, 184–192. https://doi.org/10.1016/j.agwat.2017.06.018. Dabrowska-Zielinska, K., Kogan, F., Ciolkosz, A., Gruszczynska, M., Kowalik, W., 2002. Modeling of crop growth conditions and crop yield in Poland using AVHRR based indices. Int. J. Remote Sens. 23 (6), 1109–1123. https://doi.org/10.1080/ 01431160110070744. Doorenbos, J. and Kassam, A.H., 1979. Yield response to water. FAO Irrigation and Drainage, Paper 33, Rome, 193. ISBN 978-92-5-107274-5. Farr´ e, I., Can Oijen, M., Leffelaar, P.A., Faci, J.M., 2000. Analysis of maize growth for diferent irrigation strategies in northeastern Spain. Eur. J. Agron. 12, 225–238. https://doi.org/10.1016/S1161-0301(00)00051-4. Farr´ e, I., Faci, J.-M., 2009. Deficit irirgation in maize for reducing agricultural water use in a Mediterranean environment. Agric. Water Manag. 96, 383–394. https://doi.org/ 10.1016/j.agwat.2008.07.002. Fereres, E., Orgaz, F., Gonzalez-Dugo, V., Testi, L., Villalobos, F.J., 2014. Balancing crop yield and water productivity tradeoffs in herbaceous and woody crops. Functional Plant Biology. CSIRO Publishing. https://doi.org/10.1071/FP14042. F´ eret, J.B., Gitelson, A., Noble, S., Jacquemoud, S., 2017. PROSPECT-D: towards modeling leaf optical properties through a complete lifecycle. Remote Sens. Environ. 193, 204–215. https://doi.org/10.1016/j.rse.2017.03.004. Gao, F., Kustas, W.P., Anderson, M.C., 2012. A data mining approach for sharpening thermal satellite imagery over land. Remote Sens. 4, 11. https://doi.org/10.3390/ rs4113287. Gen´ e-Mola, J., Pamies-Sans, M., Casadesús, J., Bellvert, J., 2024. Early crop type classification and mapping using sparse and multi-year remote sensing data. 2024 IEEE International Workshop on Metrology for Agriculture and Forestry, Padova (Italy), October 29-30, 445-447. ISBN: 979-8-3503-5543-7. Index | IEEE MetroAgriFor 2024. Ghrab, M., Zitouna, R., Ben Mimoun, M., Masmoudi, M.M., Ben Mechlia, N., 2013. Yield and water productivity of peach trees under continuous deficit irrigation and high evaporative demand. Biol. Agric. Hortic. 29 (1), 29–37. https://doi.org/10.1080/ 01448765.2013.750077. Girona, J., Behboudian, M.H., Mata, M., del Campo, J., Marsal, J., 2010. Exploring six reduced irrigation options under water shortage for ‘Golden Smoothee’ apple: Responses of yield components over three years. Agric. Water Manag. 98 (2), 370–375. https://doi.org/10.1016/j.agwat.2010.09.011. Girona, J., Mata, M., Marsal, J., 2005. Regulated deficit irrigation during the kernelfilling period and optimal irrigation rates in almond. Agric. Water Manag. 75 (2), 152–167. https://doi.org/10.1016/j.agwat.2004.12.008. Girona, J., Mata, M., Arbon´ es, A., Alegre, S., Rufat, J., Marsal, J., 2003. Peach tree response to single and combined regulated deficit irrigation strategies under shallow soils. J. Am. Soc. Hort. Sci. 128 (3), 432–440. https://doi.org/10.1016/j. agwat.2004.09.011. Goldhamer, D.A., Fereres, E., 2017. Establishing an almond water production function for California using long-term yield response to variable irrigation. Irrig. Sci. 35, 169–179. https://doi.org/10.1007/s00271-016-0528-2. Greaves, G.E., Wang, Y.M., 2017. Yield response, water productivity, and seasonal water production functions for maize under deficit irrigation water management in Southern Taiwan. Plant Prod. Sci. 20, 353–365. https://doi.org/10.1080/ 1343943X.2017.1365613. Guzinski, R., Nieto, H., S´ anchez, J.M., et al., 2021. Utility of copernicus-based inputs for actual evapotranspiration modeling in support of sustainable water use in agriculture. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 14, 11466–11484. https://doi.org/10.1109/JSTARS.2021.3122573. Guzinski, R., Nieto, H., Sandholt, I., et al., 2020. Modelling high-resolution actual evapotranspiration through sentinel-2 and sentinel-3 data fusion. Remote Sens. 12 (9), 1433. https://doi.org/10.3390/rs12091433. Hanks, R.J., 1974. A model for predicting plant yield as influenced by water use. Agron. J. 66 (5), 660–664. https://doi.org/10.2134/ agronj1974.00021962006600050017x. Hersbach, H., et al., 2020. The ERA5 global reanalysis, Quart. J. Roy. Meteorological Soc.146, 730, 1999–2049, [Online]. Available: 〈https://rmets.onlinelibrary.wiley. com/doi/abs/10.1002/qj.3803〉. Hoogenboom, G., C.H. Porter, V. Shelia, K.J. Boote, U. Singh, W. Pavan, F.A.A. Oliveira, L.P. Moreno-Cadena, T.B. Ferreira, J.W. White, J.I. Lizaso, D.N.L. Pequeno, B.A. Kimball, P.D. Alderman, K.R. Thorp, S.V. Cuadra, M.S. Vianna, F.J. Villalobos, W.D. Batchelor, S. Asseng, M.R. Jones, A. Hopf, H.B. Dias, L.A. Hunt, and J.W. Jones., 2023. Decision Support System for Agrotechnology Transfer (DSSAT) Version 4.8.2 (www.DSSAT.net). DSSAT Foundation, Gainesville, Florida, USA. Huang, J., Wang, H., Dai, Q., Han, D., 2014. Analysis of NDVI data for crop identification and yield estimation. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 7 (11), 4374–4384. https://doi.org/10.1109/jstars.2014.2334332. Huffman, T., Qian, B., de Jong, R., Liu, J., Wang, H., McConkey, B., Bierley, T., Yang, J., 2015. Upscaling modelled crop yields to regional scale: a case study using DSSAT for spring wheat on the Canadian prairies. Can. J. Soil Sci. 95 (1), 49–61. https://doi. org/10.4141/cjss-2014-076. Hunt, J. Van Rees, H., Hochman, Z., Carberry, P., Holzworth, D., Dalgliesh, N., Brennan, L., Poulton, P., van Rees, S., Huth, N., Peake, A., 2006. Yield prophet®: an online crop simulation service. In Proc. 13th Australian Agronomy Conference 10–14. J. Bellvert et al. Agricultural Water Management 309 (2025) 109319 19 IEMed, 2017. Mediterranean Yearbook. Water Challenges in the Mediterranean. Strategic Sectors, Economy & Territory. Panorama 274-277 [Online] Available: Water-Challenges-in-the-Mediterranean.pdf. Ines, A.V.M., Das, N.N., Hansen, J.W., 2013. Assimilation of remotely sensed soil moisture and vegetation with a crop simulation model for maize yield prediction. Remote Sens. Environ. 138, 149–164. https://doi.org/10.1016/j.rse.2013.07.018. IPCC)., 2022. Mediterranean Region. In: Climate Change 2022 – Impacts, Adaptation and Vulnerability: Working Group II Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambrigde University Press; 2023: 2233-2272. Iqbal, M.A., Shen, Y., Stricevic, R., Pei, H., Sun, H., Amiri, E., Penas, A., del Rio, S., 2014. Evaluation of the FAO AquaCrop model for winter wheat on the North China Plain under deficit irrigation from field experiment to regional yield simulation. Agric. Water Manag. 135, 61–72. Elsevier BV. http://doi.org/10.1016/j. agwat.2013.12.012. Ishaque, W., Osman, R., Hafiza, B.S., Malghani, S., Zhao, B., Xu, M., Ata-Ul-Karim, S.T., 2023. Quantifying the impacts of climate change on wheat phenology, yield, and evapotranspiration under irrigated and rainfed conditions. Agric. Water Manag. 275, 108017. https://doi.org/10.1016/j.agwat.2022.108017. Jofre-ˇ Cekalovi´ c, C., Nieto, H., Girona, J., Pamies-Sans, M., Bellvert, J., 2022. Accounting for almond crop water use under different irrigation regimes with a two-source energy balance model and copernicus-based inputs. Remote Sens. 14, 2106. https:// doi.org/10.3390/rs14092106. Johnson, D.M., 2016. A comprehensive assessment of the correlations between field crop yields and commonly used MODIS products. Int. J. Appl. Earth Obs. Geoinf. 52, 65–81. https://doi.org/10.1016/j.jag.2016.05.010. Jones, J.W., et al., 2003. The dssat cropping system model. Eur. J. Agron. 18, 235–265. Jureˇ cka, F., Fischer, M., Hlavinka, P., Balek, J., Semer´ adov´ a, D., Bl´ ahov´ a, M., Anderson, M.C., Hain, C., ˇ Zalud, Z., Trnka, M., 2021. Potential of water balance and remote sensing-based evapotranspiration models to predict yields of spring barley and winter whet in the Czech Republic. Agric. Water Manag. 256, 107064. https:// doi.org/10.1016/j.agwat.2021.107064. Keating, B.A., et al., 2003. An overview of apsim, a model designed for farming systems simulation. Eur. J. Agron. 18, 267–288. https://doi.org/10.1016/S1161-0301(02) 00108-9. Kern, A., Barcza, Z., Marjanovi´ c, ´ Arend´ as, T., Fodor, N., B´ onis, P., Bogn´ ar, P., Lichtenberg, J., 2018. Statistical modelling of crop yield in Central Europe using climate data and remote sensing vegetation indices. Agric. For. Meteorol. 260-261, 300–320. https://doi.org/10.1016/j.agrformet.2018.06.009. Kustas, W.P., Norman, J.M., 1999. Evaluation of soil and vegetation heat flux predictions using a simple two-source model with radiometric temperatures for partial canopy cover. Agric. For. Meteorol. 94, 13–29. https://doi.org/10.1016/S0168-1923(99) 00005-2. La Moncloa., 2023. 〈https://www.lamoncloa.gob.es/lang/en/gobierno/news/pagina s/2023/20230627_eu-agri-fish-council.aspx〉. Letseku, V., Grov´ e, B., 2022. Crop water productivity, applied water productivity and economic decision making. Water 14, 1598. https://doi.org/10.3390/w14101598. Li, S.H., Huguet, J.G., Schoch, P.G., Orlando, P., 1989. Response of peach tree growth and cropping to soil water deficit at various phenological stages of fruit development. J. Hortic. Sci. 64 (5), 541–552. https://doi.org/10.1080/ 14620316.1989.11515989. Li, F., Miao, Y., Chen, X., Sun, Z., Stueve, K., Yuan, F., 2022. In-season prediction of corn grain yield through PlanetScope and Sentinel-2 images. Agronomy 12, 3176. https:// doi.org/10.3390/agronomy12123176. Li, H., Porth, L., Seng Tan, K., Zhu, W., 2021. Improved index Insurance design and yield estimation using a dynamic factor forecasting approach. Insur. Math. Econ. 96, 208–221. https://doi.org/10.1016/j.insmatheco.2020.11.003. Li, Y., Zhou, Q., Zhou, J., Zhang, G., Vhen, C., Wang, J., 2014. Assimilating remote sensing information into a coupled hydrology-crop growth model to estimate regional maize yield in arid regions. Ecol. Model. 291 (10), 15–27. https://doi.org/ 10.1016/j.ecolmodel.2014.07.013. Lobell, D.B., Gourdji, S.M., 2012. The influence of climate change on global crop productivity. Plant Physiol. 160, 1686–1697. https://doi.org/10.1104/ pp.112.208298. Lopez, G., Behboudian, M.H., Vallverdu, X., Mata, M., Girona, J., Marsal, J., 2010. Mitigation of severe water stress by fruit thinning in ‘O’Henry peach: implications for fruit quality. Sci. Hortic. 125 (2), 294–300. https://doi.org/10.1016/j. scienta.2010.04.003. L´ opez-L´ opez, M., Espadafor, M., Testi, L., et al., 2018. Yield response of almond trees to transpiration deficits. Irrig. Sci. 36, 111–120. https://doi.org/10.1007/s00271-0180568-x. Lorite, I.J., García-Vila, M., Santos, C., Ruiz-Ramos, M., Fereres, E., 2013. AquaData and AquaGIS: two computer utilities for temporal and spatial simulations of waterlimited yield with AquaCrop. Comput. Electron. Agric. 96, 227–237. https://doi.org/ 10.1016/j.compag.2013.05.010. Lorite, I.J., Mateos, L., Orgaz, F., Fereres, E., 2007. Assessing deficit irrigation strategies at the leval of an irrigation district. Agric. Water Manag. 91, 51–60. https://doi.org/ 10.1016/j.agwat.2007.04.005. Lüke, A., Hack, J., 2017. Modelling hydrological ecosystem services–a state of the art model comparison. Hydrol. Earth Syst. Sci. Discuss. 1-29, 2017. https://doi.org/ 10.5194/hess-2017-436. Mahhou, A., DeJong, T.M., Shackel, K.S., Gao, T., 2006. Water stress and crop load effects on yield and fruit quality of Elegant Lady Peach [Prunus persica (L.) Batch]. Fruits 61, 407–418. https://doi.org/10.1051/fruits:2006040. MedECC., 2020. Climate and Environmental Change in the Mediterranean Basin – Current Situation and Risks for the Future. First Mediterranean Assessment Report [Cramer, W., Guiot, J., Marini, K. (eds.)] Union for the Mediterranean, Plan Bleu, UNEP/MAP, Marseille, France, 632pp. ISBN: 978-2-9577416-0-1 / DOI: 10.5281/ zenodo.7224821. Mir´ as-Avalos, J.M., Alcobendas, R., Alarc´ on, J.J., Valsesia, P., G´ enard, M., Nicol´ as, E., 2013. Assessment of the water stress effects on peach fruit quality and size using a fruit tree model, QualiTree. Agric. Water Manag. 128, 1–12. https://doi.org/ 10.1016/j.agwat.2013.06.008. Mir´ as-Avalos, J.M., Gonzalez-Dugo, V., García-Tejero, I.F., L´ opez-Urrea, R., Intrigliolo, D.S., Egea, G., 2023. Quantitative analysis of almond yield response to irrigation regimes in Mediterranean Spain. Agric. Water Manag. 279, 108208. https://doi.org/10.1016/j.agwat.2023.108208. Mu˜ noz Sabater, J., 2019. ERA5-Land hourly data from 1950 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds. e2161bac (Accessed on 19th January 2024).. Nangia, V., Oweis, T., 2016. Supplemental irrigation: a promising climate-resilience practice for sustainable dryland agriculture. In: Farooq, M., Siddique, K. (Eds.), Innovations in Dryland Agriculture. Springer, Cham. https://doi.org/10.1007/9783-319-47928-6_20. Naor, A., Girona, J., 2012. Fruit trees and vines: apple. In: Crop yield response to water. Steduto, P., Hsiao, T.C., Fereres, E., Raes, D. (Ed.). FAO Irrigation and Drainage Paper 66: 332-345. Norman, J.M., Kustas, W., Humes, K., 1995. A two-source approach for estimating soil and vegetation energy fluxes from observations of directional radiometric surface temperature. Agric. For. Meteorol. 77, 263–293. https://doi.org/10.1016/01681923(95)02265-Y. Olivo, N., Girona, J., Marsal, J., 2009. Seasonal sensitivity of stem water potential to vapour pressure deficit in grapevine. Irrig. Sci. 27, 175–182. https://doi.org/ 10.1007/s00271-008-0134-z. Paolini, G., Escorihuela, M.J., Merlin, O., Pamies-Sans, M., Bellvert, J., 2022. Classification of different irrigation systems at field scale using time-series of remote sensing data. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 15, 10055–10072. https://doi.org/10.1109/JSTARS.2022.3222884. Payero, J.O., Tarkalson, D.D., Irmak, S., Davison, D., Petersen, J.L., 2008. Effect of irrigation amounts applied with subsurface drip irrigation on corn evapotranspiration, yield, water use efficiency, and dry matter production in a semiarid climate. Agric. Water Manag. 95 (8), 895–908. https://doi.org/10.1016/j. agwat.2008.02.015. Prasad, A.K., Chai, L., Singh, R.P., kafatos, M., 2006. Crop yield estimation model for Iowa using remote sensing and Surface parametres. Int. J. Appl. Earth Obs. Geoinf. 8 (1), 26–33. https://doi.org/10.1016/j.jag.2005.06.002. Priestley, C.H.B., Taylor, R.J., 1972. On the assessment of surface heat flux and evaporation using large-scale parameters. Mon. Weather Rev. 100, 81–92 https:// doi.org/10.1175/1520-0493(1972)100<0081:OTAOSH>2.3.CO;2. Rocha, J., Carvalho-Santos, C., Diogo, P., Beça, P., Keizer, J.J., Nunes, J.P., 2020. Impacts of climate change on reservoir water availability, quality and irrigation needs in a water scarce Mediterranean region (Southern Portugal. Sci. Total Environ. 736, 139477. https://doi.org/10.1016/j.scitotenv.2020.139477. Romero, P., Navarro, J.-M., García, F., Ordaz, P.B., 2004. Effects of regulated deficit irrigation during the pre-harvest period on gas exchange, leaf development and crop yield of mature almond trees. Tree Physiol. 24, 303–312. https://doi.org/10.1093/ treephys/24.3.303. Rußwurm, M., K¨ orner, M., 2018. Multi-temporal land cover classification with sequential recurrent encoders. ISPRS Int. J. Geo Inf. 7, 129. https://doi.org/10.3390/ ijgi7040129. Steduto, T.C., Hsiao, E., Fereres., Raes, D., 2012. Crop Yield Response to Water. FAO Irrigation and Drainage Paper 66. By P. Rome, Italy: Food and Agriculture Organization of the United Nations, 500, ISBN 978-92-5-107274-5. Steduto, P., Hsiao, T.C., Raes, D., Fereres, E., 2009. Aquacrop-the FAO crop model to simulate yield response to water: I. Concepts and underlying principles. Agron. J. 101 (3), 426–437. https://doi.org/10.2134/agronj2008.0139s. Stockle, C.O. and Nelson, R., 1994. CropSyst User’s Manual. Biological Systems Engineering Dept., Pullman, Washington, USA, Washington State University. Teal, R.K., Tubana, B., Girma, K., Freeman, K.W., Arnall, D.B., Walsh, O., Raun, W.R., 2006. In-season prediction of corn grain yield potential using normalized difference vegetation index. Agron. J. 98 (6), 1488–1494. https://doi.org/10.2134/ agronj2006.0103. Tong, X., Wu, P., Liu, X., Zhang, L., Zhou, W., Wang, Z., 2022. A global meta-analysis of fruit tree yield and water use efficiency under deficit irrigation. Agric. Water Manag. 260, 107321. https://doi.org/10.1016/j.agwat.2021.107321. Vazifedoust, M., van Dam, J.C., Bastiaanssen, W.G.M., Feddes, R.A., 2009. Assimilation of satellite data into agrohydrological models to improve crop yield forecasts. Int. J. Remote Sens. 30 (10), 2523–2545. https://doi.org/10.1080/01431160802552769. WorldBank., 2023. https://www.worldbank.org/en/topic/water-in-agriculture#1. Yang, Y., Anderson, M.C., Gao, F., Wardlow, B., Hain, C.R., Otkin, J.A., Alfieri, J., Yang, Y., Sun, L., Dulaney, W., 2018. Field-scale mapping of evaporative stress indicators of crop yield: an application over Mead, NE, USA. Remote Sens. Environ. 210, 387–402. https://doi.org/10.1016/j.rse.2018.02.020. Yang, Y., Anderson, M.C., Gao, F., Johnson, D.M., Yang, Y., Sun, L., Dulaney, W., Hain, C. R., Otkin, J.A., Prueger, J., Meyers, T.P., Bernacchi, C.J., Moore, C.E., 2021. Phenological corrections to a field-scale, ET-based crop stress indicator: an application to yield forecasting across the U.S. Corn Belt. Remote Sens. Environ. 257, 112337. https://doi.org/10.1016/j.rse.2021.112337. J. Bellvert et al.